PROCESS FOR THE HYDROGENATION OF HOP RESIN ACIDS

Description

[0001] This invention relates to the production of hydrogenated forms of hop resin acids, useful for the bittering or foam enhancement of beer. (By hop resin acids we mean the α-acids or β-acids extractable from dried hop cones of the plant Humulus lupulus, the isomerised derivatives of α-acids (especially iso-α-acids), and any reduced forms of the aforesaid hop resin acids capable of further hydrogenation by our process (especially including rho-iso-α-acids) and any mixture thereof.). Particularly, it relates to the production of hydrogenated forms of isomerised α-acids, especially tetrahydroiso-α-acids and hexahydroiso-α-acids.

[0002] The bitter flavour of beer produced in a traditional fashion is known to derive in the most part from the naturally occurring α-acids (primarily cohumulone, humulone and adhumulone) that are present in the lupulin glands of hop cones. In the traditional brewing process, dried hop cones are added to "sweet" wort and boiled for at least an hour in the brewery kettle. This boiling process extracts the α-acids and converts a portion to the corresponding isomeric forms known as iso-α-acids (primarily isocohumulone, isohumulone and isoadhumulone). It is these latter substances that constitute the major bittering substances present in the beer following fermentation of the boiled wort by yeast. They are also known to be vital components of beer foam. Unfortunately, the utilisation of hop substances in the traditional brewing process is poor, and it is typical for only about 30% of the α-acids present in the hops to appear as iso-α-acids in the beer. However, it was long ago determined that the separate extraction of the α-acids from hops by a non-polar solvent, followed by their chemically induced conversion to iso-α-acids, offered the possibility to greatly increase the efficiency by which beer can be bittered with hops. Nowadays, in a typical, modem process, hops are first extracted with liquid or supercritical CO₂. The resultant extract contains α-acids, β-acids (primarily colupulone, lupulone and adlupulone), hop oils and some low molecular weight fats and waxes. The α-acids present in the extract may then be converted to iso-α-acids by the application of heat and alkali in an aqueous suspension, typically in the presence of magnesium ions, which are known to accelerate the isomerisation reaction (see US Patent No. 5,370,897 to Smith & Wilson). The iso-α-acids so formed may then be separated from the reaction mixture and purified, typically through fractional precipitation steps using a mineral acid to reduce the pH value of the aqueous phase. For commerciai purposes, it is usual for the purified iso-α-acids to be converted to the potassium salt form and standardised as a clear, slightly alkaline, aqueous solution at a concentration of about 30%. This solution may then conveniently be used by brewers to directly bitter worts that have already been fermented. The use of such iso-α-acids preparations in this way can improve the utilisation of the original hop α-acids to as much as 80% and, consequently, in most cases is found to substantially reduce the cost of bittering beer.

[0003] Although the addition of an iso-α-acids preparation to beer is commonplace, its advantages do not meet the requirements of all brewers. For instance, those brewers who wish to market beer containing iso-α-acids in clear or green glass bottles risk the development of so-called "lightstruck", "sunstruck" or "skunky" flavour if the beer is exposed to sunlight or to certain types of artificial light. This flavour is known to be due to the formation of 3-methyl-2-butene-1-thiol (MBT) via a photolytic reaction between iso-α-acids and naturally occurring sulphur compounds in the beer. (See review by Templar, Arrigan & Simpson in Brewers Digest, May 1995, p18-25). MBT has an exceptionally low taste threshold (variously estimated as between about 2 and 300 parts per trillion in beer, according to the nature of the beer and the sensitivity of the individual taster) and its presence is generally considered to be undesirable. The possibility for the occurrence of the photolytic reaction is known to be dependent upon the presence of an isohexenoyl side chain of the iso-α-acids. If the molecular structure of this side chain is modified, for example by the reduction of either the carbon-carbon double bond or the carbonyl group, then MBT can no longer be produced. (Presumably, this is because it is no longer possible to form the 3-methylbutenyl radical that is believed to be the precursor of MBT, forming this objectionable substance via condensation with a free thiol radical). Consequently, various reduced derivatives of the iso-α-acids have been prepared and their properties described. All are bitter, though the intensities of bitterness produced when these compounds are added to beer are demonstrably different. However, in each case, when these compounds are used as the sole source of bittering in a beer, whether independently or as a mixture, it is invariably found that the beer is effectively protected from the formation of lightstruck flavour, providing only that there is no contamination by non-reduced iso-α-acids.

[0004] Three reduced derivatives of the iso-α-acids are now commercially available and in use by many brewers. (For a comprehensive summary of methods of manufacture, properties and uses, see European Brewery Convention Manual of Good Practice, "Hops and Hop Products ", (1997), Section 5.4.4 (p 93 - 106), pub. by Getränke-Fachverlag Hans Carl, Nürnberg, Germany). Specifically, these are: the rho-iso-α-acids, (a dihydroiso-α-acids derivative characterised by reduction of the aforementioned side chain carbonyl group (-C=O) to an hydroxyl group (-CH(OH)); the tetrahydroiso-α-acids (characterised by reduction of the aforementioned carbon-carbon double bond (-CH=CH-) to a single bond (-CH₂-CH₂-) and also by reduction of a carbon-carbon double bond in another, isopentenyl side chain) and the hexahydroiso-α-acids (characterised by reduction of both of the aforementioned carbon-carbon double bonds and the carbonyl group of the isohexenoyl side chain). The properties of these three derivatives are such as to allow the brewer not only to eliminate the possibility of forming lightstruck flavour, but also to usefully modify the foam stability of his beer. Relative to iso-α-acids, rho-iso-α-acids are relatively low in bitterness and have somewhat reduced foam stabilising properties. On the other hand, tetrahydroiso-α-acids are relatively more bitter and their use enhances beer foam stability, even at equivalent bittering level. Hexahydroiso-α-acids are exceptionally foam enhancing but are considered to have a bittering effect that is only a little greater than that of iso-α-acids. Clearly, by judicious use of these compounds, the brewer is able to manipulate two of the important characteristics of beer so as to produce a desired combination of qualities. Hence, it follows that the use of these reduced iso-α-acids is not necessarily limited only to those brewers who seek to produce lightstable beers. Indeed, such is the perceived value of these compounds that the major market for the tetrahydroiso- and hexahydroiso-α-acids is probably created by their primary use as foam enhancing additives to beers that already contain iso-α-acids deriving from conventional addition of hops, hop pellets or extracts to the wort kettle.

[0005] Several methods have been described for the production of the reduced iso-α-acids. In the case of the rho-iso-α-acids, all published procedures achieve this product via borohydride mediated reduction of iso-α-acids. For example, Westermann et al (US Patent No. 3,558,326) showed how the α-acids content of a concentrated hexane extract of hops could be simultaneously isomerised and reduced to form rho-iso-α-acids in the presence of a hot, aqueous, strongly alkaline solution of sodium borohydride. After separation of the hexane to remove hop oils and waxes, the aqueous phase was contacted with fresh hexane and then acidified with sufficient sulphuric acid to force the reduced (rho-)iso-α-acids into the organic solvent phase. The latter phase was then recovered by centrifugation and the hexane removed by evaporation to leave a product containing a mixture of rho-iso-α-acids and unreacted β-acids. In an alternative process, Goldstein et al (US Patent No. 4,324,810) teach the production of a similar product, this product being prepared in a way that avoids the use of organic solvents (only excepting any such solvent that may have been used to prepare the hop extract that was the starting point for their process). Again, the isomerisation and reduction of the hop α-acids was achieved by dissolution and heating in an alkaline, aqueous solution of sodium borohydride. Such reduction by sodium (or potassium) borohydride of the side chain carbonyl group is very specific and occurs without significant reduction of the carbon-carbon double bonds.

[0006] Tetrahydroiso-α-acids have been produced by several, substantially different processes and from different hop resin acid substrates. Worden & Todd (US Patent No. 3,552,975) describe a process for the formation of desoxytetrahydro-α-acids via catalytic hydrogenolysis and hydrogenation of β-acids (a by-product of the production of iso-α-acids) using hydrogen gas and a palladium on carbon or other noble metal catalyst. These intermediates are then subjected to a process of oxidation by peracetic acid, followed by isomerisation of the resultant tetrahydro-α-acids to form the tetrahydroiso-α-acids. Their process uses substantial quantities of organic solvents and risks leaving undesirable residues of these solvents in the final product. Cowles et al (US Patent No. 4,644,084) improve upon this process by oxidising the desoxytetrahydro-α-acids in alkaline, aqueous alcohol with an oxygen containing gas stream, but this variant method is still complex and also requires the use of substantial amounts of ethanol. However, it does also achieve isomerisation of the intermediate tetrahydro-α-acids in the same process step. The formation of tetrahydroiso-α-acids from α-acids, rather than from the less valuable β-acids, is an inherently easier task. For example, Verzele & Anteunis (in Bull. Soc. Chim. Belg., Vol. 68, (1959), p 315 - 324) described a simple, laboratory scale process in which humulone, previously separated from a hop extract, was hydrogenated in methanol to tetrahydrohumulone by exposure to hydrogen gas in the presence of a platinum oxide catalyst. Brown, Howard & Tatchell (J. Chem. Soc. (1959), p 545 - 551) separately showed how such a tetrahydro-α-acid (in their case, tetrahydrocohumulone,) could then be isomerised by application of heat and ethanolic alkali to form the corresponding tetrahydroiso-α-acid. The alternative possibility of first isomerising an α-acid and then subjecting it to catalytic reduction by hydrogen gas in an organic solvent to produce a tetrahydroiso-α-acid derivative was also described in this paper. However, the best yield reported for the hydrogenation (of isohumulone) was only 70%, and it was clear from this work that there was a tendency either for incomplete reaction (leading to substantial formation of dihydroisohumulone) or perhydrogenation (as indicated by the formation of neo-hydroisohumulone). Indeed, according to Hay & Homiski (in J. Agric. Food Chem. Vol. 39, (1991), p 1732-1734) the prior art up until the time of writing had indicated that production of tetrahydroiso-α-acids in good yield by direct hydrogenation from pre-isomerised α-acids had never been achieved. Nevertheless, Hay demonstrated (US Patent 5,013,571) that an efficient conversion was, in fact, possible and could be achieved over a Pd/C catalyst, whether in aqueous, aqueous alcoholic or chlorinated hydrocarbon solution. Poyner et al (US Patent No. 5,600,012) also teach the direct hydrogenation of iso-α-acids, restricting their process to dissolution in pure ethanol and in the presence of specific catalysts based on palladium oxide (as opposed to palladium metal) that were not prone to induce perhydrogenation of the substrate. Importantly, Hay also claimed in his patent that it was possible to achieve both the isomerisation and the hydrogenation in a single step by exposing the α-acids to hydrogen and noble metal catalyst in such a solution, provided that the temperature was above about 50°C. Thus, Hay was able to demonstrate for the first time the possibility to form tetrahydroiso-α-acids in a purely aqueous environment, though he failed to demonstrate the subsequent separation of this product from the reaction mixture without the aid of added alcohol.

[0007] The earlier work of Anteunis & Verzele (in Bull. Soc. Chim. Belg., 68, 456 - 583 (1959)) showed that the direct reduction of α-acids to tetrahydro-α-acids, especially under acidic conditions, was accompanied by the formation through hydrogenolysis of substantial amounts of humulohydroquinones, these being substances of no known value in brewing. Subsequently, Stegink, Guzinski & Todd (US Patent No. 5,296,637) claimed that the occurrence of hydrogenolysis and the difficulty of preventing perhydrogenation had ensured that "the art has not been able to utilize α-acids as a source of tetrahydrohumulates, with subsequent isomerization to tetraisohumulates". However, they were able to demonstrate that these undesirable reactions could be virtually eliminated by hydrogenating the α-acids as alkaline metal salts in aqueous or alcoholic solution where the pH value was sufficiently high to effectively ensure the absence of their natural, acidic form. Having achieved the successful hydrogenation of the α-acids, Stegink et al noted that, if desired, tetrahydroiso-α-acids could then easily be formed by boiling the tetrahydro-α-acids in aqueous solution at pH 10. Stegink et al also claimed to have demonstrated that their process enabled the production of tetrahydroiso-α-acids having superior purity and in higher yield than was at that time obtained commercially from hydrogenation of iso-α-acids, presumably by application of the Hay process. Particularly, they claimed that hydrogenation via their method ensured the absence of perhydrogenation reactions that they demonstrated were liable to occur under the conditions described by Hay, a problem that it has already been noted was later solved by Poyner et al, though only by restriction of their process to the use of the palladium catalyst in a specific, oxidised form. As we shall later demonstrate, it is an advantage of our invention that this restriction to the use of palladium oxide catalysts in order to avoid perhydrogenation is, in our process, found to be unnecessary.

[0008] Two fundamental approaches have been taken to devising processes for the formation of hexahydroiso-α-acids. First described, was the use of tetrahydroiso-α-acids as the starting point wherein the desired formation of the hexahydroiso-α-acids was achieved by a borohydride reduction of the tetrahydroiso-α-acids. Thus, Worden & Todd (US Patent No. 3,552,975) described how potassium borohydride was added to an alkaline solution of tetrahydroiso-α-acids (previously prepared from β-acids) and the mixture allowed to react for several hours at room temperature. The resultant hexahydroiso-α-acids were then recovered by acidifying the solution and extracting the hop resin acids into dichloromethane. The solvent was then dried over sodium sulphate and finally evaporated away so as to obtain the hexahydroiso-α-acids as a solvent free, pale yellow gum. By processing from tetrahydroiso-α-acids, Worden and Todd were, of course, effectively also demonstrating the possibility to synthesise hexahydroiso-α-acids from either the α-acids or the β-acids of hops (though they did not in fact themselves demonstrate the complete synthesis from α-acids). Much later, Hay (US Patent No. 5,013,571) demonstrated that hexahydroiso-α-acids could also be formed from α-acids by a reversal of the necessary reduction stages. Thus, Hay describes how rho-iso-α-acids were successfully hydrogenated in ethanol over a 10% palladium on carbon catalyst. In a variant of his process, Hay also showed how an aqueous, alkaline methanolic solution of α-acids could be reduced to (primarily) hexahydroiso-α-acids in a single reaction stage by reacting the α-acids with alkaline borohydride by heating in the presence of hydrogen gas and the same 10% Pd/C catalyst.

[0009] As may be seen, the hydrogenation processes described above depend upon the use of conventional organic and/or aqueous solvents in one form or another. Clearly, the use of (protic) solvents during the hydrogenation of iso-α-acids under alkaline conditions as in the process of Hay is an absolute requirement since the iso-α-acids must be presented to the catalyst in an anionic form. Although the process of Poyner et al relies upon maintenance of the un-ionised, acidic form of the iso-α-acids, nowhere do they suggest that catalysis can be successful without the prior dissolution of the substrate into organic solvent (in their case restricted to alcohol).

[0010] In their natural, acidic state, hop resin acids are essentially insoluble in water unless the pH is raised by, for example, addition of an alkali metal hydroxide. As the acid resin is converted into salt form, so the solubility increases, reaching a maximum when the pH value exceeds the pK_a of the hop acid by at least 3 pH units. Conversely, as the pH of a concentrated solution is reduced by, for example, addition of a mineral acid such as sulphuric or hydrochloric acid, the hop acid will be progressively precipitated from the solution. Thus, for example, whilst in commercial practice iso-α-acids are typically sold as a 30% (w/v) clear, aqueous solution of their potassium salts at a pH value of between 9.0 and 10.0, reduction of the solution pH to below 4.0 by addition of a mineral acid will cause precipitation of most of the hop acid as a somewhat mobile liquid, and further reduction to below pH 2.0 will cause almost total precipitation to occur. This principle is known to be used in commercial practice during the manufacture of iso-α-acids and results in the formation of a concentrated iso-α-acids resin that is typically about 90% pure iso-α-acids, the remainder being mostly resinous, organic acid by-products. Dependent upon temperature, an incompletely purified hop resin acid, typically containing a mixture of homologues (and possible stereoisomers) and in its natural, acidic state, typically has the form of an amorphous solid, a sticky semi-solid paste or a resinous, relatively (to water) viscous liquid. In the latter situation, this phenomenon is often used as a practical means to prepare a concentrate of the hop acid since the precipitated, mobile hop resin acid typically has a different density to that of the aqueous phase and can be readily separated by gravitational means such as centrifugation or even merely by a process of settling and decantation. The temperature at which the transition from solid or semi-solid to variously mobile liquid takes place is dependent upon the exact composition of the hop resin acid. Different, pure isomers of closely related, individual homologues of hop resin acids may have quite different properties in this respect. For example, Verzele and De Keukeleire (in Chemistry and Analysis of Hop and Beer Bitter Acids, Elsevier Science Publishers B.V., Amsterdam/London/New York/Tokyo, 1991, pages 92, 93, 129 and 135) found that, whilst cis-isohumulone had a melting point of 18°C, its diastereoisomer, trans-isohumulone did not melt until the temperature was raised to 65°C. Chemical reduction of these compounds normally raises the melting point: thus, trans-tetrahydroiso-α-acid reportedly melts at 75.5 - 78°C and one of the two possible stereoisomers of trans-rho-iso-α-acid, at 80°C. Since the α-acids fraction of hops is invariably found to contain three major and several minor homologues, and their isomerised derivatives have more than one stereoisomeric form, it is only to be expected that, for example, the physical characteristics, in particular the viscosity of different isomerised α-acid resin preparations containing mixtures of homologues and isomers, will be seen to vary according to the variety of hops from which they are made and the processes used for their preparation.

[0011] Naturally, the direct hydrogenation of a relatively viscous hop resin acid in its natural, acidic form may be expected to be rather slow and particularly difficult to take to completion. It is known that the rate limiting step in certain hydrogenation reactions is the rate of transfer of substrate(s) (including hydrogen) and product(s) across the boundary layer surrounding the catalyst particles, a parameter that is inversely related to the viscosity of the medium. A similar situation exists within the pores of the catalyst, where the rate of diffusion will also be inversely related to the fluid viscosity. Where the medium is a viscous substrate itself, the reaction may perhaps start at a reasonable rate due to the high substrate concentration, but could be expected at some stage to become progressively slower as the medium itself is converted into the (normally) still viscous product. More commonly in industrial hydrogenation processes though, the over-riding limitation on the achievable reaction rate is the inadequate availability of hydrogen gas to the active sites on the surface and within the pores of the catalyst particles. Such limitation can be due to inherent low solubility and poor dissolution of the hydrogen gas into the liquid medium resulting in a rate of transfer through the boundary layer and within the pores of the catalyst that fails to match the capacity of the active sites to reduce the available substrate. In such cases, it is often found that the intended reaction is partially or wholly replaced by isomerisation reactions that are usually favoured by conditions of hydrogen starvation. The use of special, often expensive, "eggshell" catalysts (where the active sites of the catalyst are restricted to a thin layer surrounding the matrix of the catalyst particle itself) is often recommended as a means to avoid such a situation developing during a hydrogenation. Various, often proprietary methods to induce an extremely vigorous mixing of the hydrogen gas with the liquid medium are commonly employed in another, non-exclusive strategy to maximise the rate at which hydrogen is dissolved, thereby increasing the driving force for the transfer of hydrogen molecules through the boundary layer. This approach to improving hydrogen transfer rates is particularly necessary when a hydrogenation reaction is carried out in water, the preferred solvent of the aforementioned Hay process for conversion of iso-α-acids to tetrahydroiso-α-acids, since the solubility of hydrogen in water is quite low (Braker & Mossman in Matheson Gas Data Book (Fifth Edition, 1971, Matheson Gas Products, East Rutherford, NJ, USA), quote 0.019 vol/vol at 60°F, I atm.). In such a situation, reactions can be accelerated (and the possibility of unwanted alternate reactions reduced) by increasing the pressure of the hydrogen. However, such a course of action not only requires the provision of a more expensive reaction autoclave, but may also have the disadvantage that the reaction is harder to monitor and control because of the excess of hydrogen necessarily supplied and the relatively small change in the measured gas pressure. In the case of the hydrogenation of hop resin acids, this inevitably increases the possibility of inducing undesirable perhydrogenation of the substrate. The higher pressure of hydrogen also increases the potential for encountering a dangerous leak of highly flammable gas and the creation of a potentially explosive atmosphere in the immediate vicinity of the autoclave.

[0012] One solution to the problem of insufficient or poor rate of transfer of either or both of the hydrogen and substrate molecules to the active sites of the catalyst has been described by Poliakoff et al in PCT International Publication No. WO 97/38955. These authors teach a process for the hydrogenation of a wide range of functional groups in aliphatic and aromatic organic compounds in which both substrate and hydrogen are dissolved at low, but readily controllable concentrations into a continuously flowing stream of carrier fluid that is under supercritical or near-critical conditions, the mixture so formed then being passed through a sufficient bed of catalyst that the desired hydrogenation reaction is effectively completed during passage of the reaction mixture across the catalyst particles. The product of the reaction can easily be recovered by reducing the pressure of the fluid exiting the reactor sufficiently that the carrier substance is converted into the gaseous state, thereby precipitating the product as a separate liquid or, possibly, solid phase. The two phases may then be discharged into a collection vessel which allows for the retention of the product and discharge of the gaseous phase (including unreacted hydrogen) to atmosphere or into a suitable recycling system. Several carrier fluids are claimed as suitable for the operation of this process, but particularly mentioned are propane and carbon dioxide, the latter substance being especially favoured because of its generally inert and non-flammable nature, thus ensuring that any leakage of fluid from the process is necessarily rendered non-flammable and non-explosive. In liquid form, both of these substances are also known to have very low absolute viscosities, especially when under supercritical conditions rather than as liquids at temperatures below their critical temperatures. Excepting at very high pressures (relative to the critical pressure) their density is somewhat reduced, too. Hence, the rate of mass transfer of solutes in the supercritical medium may be expected to be much faster than is the case in conventional (i.e. non-supercritical) hydrogenation processes of the prior art. And although Poliakoff et al note in the preamble to their application that there had been previous attempts to hydrogenate substances in supercritical fluids, they dismiss these as unsatisfactory, stating (page 4, lines 22 - 27) that "Accordingly, there is a need for a hydrogenation process which can be effected under conditions such that only small quantities of the organic compound and hydrogen are required in the reactor at any one time, and yet which can be used on an industrial scale for the manufacture of hydrogenated products". Thus, whilst on the one hand Poliakoff et al reportedly solve the problem of mass transfer and, additionally, demonstrate that they readily can control and manipulate the stoichiometry of many possible hydrogenation reactions by varying key parameters such as temperature and the ratio of hydrogen to substrate, they necessarily complicate matters by the requirement to operate their process with a large excess of the carrier fluid.

[0013] Subsequent to the work of Poliakoff et al, the further teachings of Hill, Muller and Swidersky (UK Patent Application No. GB 2,336,363) demonstrate by their examples how hop (resin) acids may successfully be hydrogenated by contacting the hop resin acid with hydrogen gas (at a partial pressure of 20bar) and a palladium on carbon catalyst in the presence of supercritical CO₂ (i.e. at a temperature above 31°C and a pressure in excess of 73 bar). In discussing the working of their process, Hill et al state (page 2) "Normally, the pressure will not exceed 1000 bar, and is typically in the range of 200 to 350 bar" and also add "Preferably, the temperature is above 40°C and typically around 75°C". However, despite the use of high pressures, and temperatures significantly above ambient, Hill et al note (page 3): "The hop acids are typically in solution at a loading of about 2.5 - 5% wt/vol.", thereby confirming a necessity to work with a relatively large excess of solvating fluid.

[0014] It is an object of our invention to demonstrate that, most surprisingly, the need to use conventional solvents, including water and organic liquids such as alcohols or hydrocarbons, as carriers for the hydrogenation of iso-α-acids and rho-iso-α-acids in a necessarily dissolved state may be eliminated. Thus, contrary to natural expectations and the implications of prior art, hydrogenation of iso-α-acids and other hop resin acids can be achieved entirely without the use of any conventional solvent, thereby eliminating the necessity for incorporation in a process of a step for subsequent removal of said solvent.

[0015] A more particular object of our invention is to provide a process whereby iso-α-acids resin (or rho-iso-α-acids resin) may be directly hydrogenated to form tetrahydroiso-α-acids resin (or hexahydroiso-α-acids resin) of high purity that may be readily converted into a soluble, aqueous form suitable for use by brewers desirous to prepare beers having improved foam characteristics or lightstability.

[0016] Our invention provides a process whereby a specific hop resin acid (such term being taken to include mixtures of closely related homologues) or mixtures of such acids are hydrogenated by direct contact with a noble metal catalyst under an atmosphere containing hydrogen gas and in the complete absence of liquid, organic solvents.

[0017] More specifically, the process according to the present invention is a process for the hydrogenation of hop resin acids in their acidic state, and in the absence of liquid, organic solvents, by contacting a reaction mixture of said hop resin acids with hydrogen gas in the presence of a noble metal catalyst under partial pressure of carbon dioxide (CO₂), the pressure of said CO₂ gas being (a) sufficient to fluidise said reaction mixture, and (b) below that at which the said CO₂ gas will liquefy at below the critical temperature for CO₂, or at pressure less than that at which the density of the CO₂ gas exceeds its value at the critical point if the temperature is at or above the said critical temperature.

[0018] Preferred embodiments of the process according to the present invention are characterized by the additional features of appending claims 2 to 6.

[0019] Some hop resin acids are particularly viscous at normal room temperature or may even partially crystallise or form a semi-solid paste. In order to facilitate hydrogenation of such substances, our process increases the fluidity of the hop resin acid(s) substrate, thereby enabling the hydrogenation to either take place at a lower temperature than would otherwise be practical because of excessive viscosity of the hop resin acid/catalyst mixture or, alternatively, to have the effect of increasing the rate of reaction at a given temperature. This improvement is achieved by introducing CO₂ gas to the autoclave at a sufficiently high partial pressure to force the dissolution of sufficient CO₂ into the substrate as to cause the viscosity to be substantially reduced, to or beyond the point at which the degree of agitation within the resin/catalyst mixture is enough to enable the hydrogenation reaction to take place at a convenient rate and facilitated in part by an increase in the solubility of the hydrogen that we believe is caused by the presence of CO₂ in the hop resin acid. This pressure of added CO₂ gas is, typically, in excess of that which would be required just to ensure that any leakage of gas from the reactor would not be flammable or potentially explosive, but is necessarily insufficient to cause liquefaction to take place should the temperature be below the critical temperature (31.0°C for pure CO₂). Where the selected reaction temperature exceeds the critical temperature, no liquefaction is possible and the pressure may be allowed to marginally exceed the pressure at the critical temperature (7.4 MPa absolute [1072 psia]), but not to the point at which the gas density exceeds that at the critical point (under which circumstances the CO₂ may be considered to be in a supercritical condition and capable of dissolving at least a substantial portion of the hop resin acid). In other respects, the reaction conditions may be similar or even the same as for the working of an unmodified process as described above, excepting that it will now be found possible to achieve substantially the same desired effects but at an optionally lower temperature.

[0020] The hydrogenated hop resin acid product may be subsequently separated from the catalyst by, for example, filtration or centrifugation of the resin/catalyst mixture at or somewhat above whatever temperature the resinous product is found to be sufficiently fluid to facilitate the separation process. In some instances, the clarified product may then be of such a nature as to be found suitable for use by a brewer for addition to wort or beer. More typically though, it will be desired to convert this hydrogenated hop resin acid into a solubilised form, most commonly as an aqueous, alkaline solution of standardised concentration, but optionally as a solution in a food grade solvent such as propylene glycol or ethanol. A convenient, but by no means exclusive method by which the former may be achieved is by suspending the resin in water in a stirred tank at a temperature such that the hydrogenated hop resin acid is in a fluid state and adding an alkali, such as a concentrated solution of an alkali metal hydroxide, until the resin is totally dissolved. The resultant solution of hydrogenated hop resin acids may then used in whatever way the brewer desires to add flavour and/or other characteristics to wort or beer.

[0021] In an alternative method for converting the hydrogenated hop resin acid/catalyst mixture into a solubilised form, the conversion of the resinous product into a clarified and standardised solution may more advantageously be achieved by solubilising the resin before separation from the catalyst. In this variant of our process where the end product is to be an aqueous, alkaline solution, the resin/catalyst mixture may conveniently be first suspended and stirred in water at a temperature at which the resin is not excessively viscous and then dissolved by addition of a concentrated solution of an alkali hydroxide such as potassium hydroxide. The catalyst can then be more easily removed from the mixture by filtration or centrifugation than typically would be the case for separation from the hydrogenated hop resin acid in its resinous state. Sometimes, the recovery of the solubilised resin is best if the solution is deliberately made quite dilute before the catalyst is removed. In such a case, the resin may be readily reconstituted by precipitation, and subsequent collection as a separate phase, via the addition to the (stirred) solution of a concentrated mineral acid such as 50% sulphuric acid in sufficient amount as to convert the anionic form of the resin back to its acidic, non-ionised state. Separation of the phases usually will be found to take place after cessation of agitation and it is then an easy matter to recover the resin, only provided that the temperature is kept sufficiently high to maintain the resin in a fluid condition. The desired, aqueous, alkaline solution then may be reconstituted at the correct strength by suspension in water and addition of alkali as previously described in the case of a hydrogenated resin acids product formed by clarification to remove catalyst directly following the hydrogenation reaction. An inherent advantage of including in a process the steps of precipitation, separation and reconstitution of the resin is that the purity of the product may thereby be enhanced since any highly acidic, organic contaminants or by-products of the hydrogenation reaction may be discarded in the aqueous, acidic waste from the separation process.

[0022] Example 1 illustrates an embodiment of our invention that enables hydrogenation to be conducted advantageously at a reduced temperature, typically at or about ambient temperature. It was earlier noted that the problem of inadequate mass transfer has previously been solved by dissolving organic reactants into a carrier gas that is in a relatively dense, near-critical liquid or truly supercritical fluid state. Carbon dioxide in such a condition was mentioned as being a particularly suitable solvent for many organic, catalysed hydrogenation reactions, albeit that the concentration of the reactant is necessarily limited by its solubility in the carrier fluid. This example demonstrates that, surprisingly, mere contact with CO₂ in its purely gaseous state can sufficiently reduce the viscosity of an otherwise excessively viscous hop resin acid that its direct hydrogenation can subsequently proceed at a satisfactory rate:

EXAMPLE 1 (Direct Hydrogenation of Iso-α-acids Resin at Low Temperature and under High Pressure of CO₂ Gas)

[0023] 10.65 g of iso-α-acids ("Preparation I ") was placed into an autoclave, together with 0.583 g of 10% PdO/C catalyst ("Catalyst 1"). The proportion of catalyst to iso-α-acids was therefore 3.1% on a dry weight basis. The vessel was then purged of air, evacuated and then brought to an internal pressure of 0.45 MPa. (65 psig) with hydrogen gas. CO₂ gas was then admitted into the autoclave from a laboratory gas cylinder containing liquid CO2 and the pressure allowed to rise to 4.9 MPa (715 psig)before closing the gas inlet valve. The autoclave was then partially immersed into a water bath maintained at only 20°C. The contents of the vessel were magnetically stirred for 22 min, during which time the pressure fell to 4.8 MPa psig). While the contents continued to be stirred, further CO₂ gas was then admitted over a short period until the pressure had risen to 5.1 MPa (745 psig) and the inlet valve was again closed. After a further 25 min, the pressure had fallen to 4.96 MPa (720 psig) at which point it was again manually raised, this time to 5 MPa (730 psig). The stirrer was eventually stopped after a total mixing time of 2.2 h and the pressurised gas vented to atmosphere over a period of several minutes. The vessel was then opened and a sample of the contents withdrawn, diluted into methanol and filtered for subsequent analysis. As shown in Table 1, the original iso-α-acids content of the substrate resin had been efficiently converted to tetrahydroiso-α-acids having a purity (by HPLC) of 93.6% and an A_254nm/A_270nm ratio of 1.19, indicative of a low level of perhydrogenation. There was no trace of either the original iso-α-acids or of intermediate dihydroiso-α-acids, confirming that the reaction was complete.

[0024] This Example demonstrates how direct, solvent free hydrogenation of what may be presumed to be otherwise excessively viscous resin is made achievable by the application of a high pressure of CO₂ gas but without the necessity to resort to conditions in which this CO₂ gas would be converted into a liquid or supercritical state. Addition of CO₂ gas in such a way has the additional benefit of rendering the gaseous phase non-flammable and non-explosive, thereby eliminating the serious risk of fire and ignition-induced explosion that is associated with the use of pure hydrogen gas. While not wishing to be bound by theory, it is believed that, under relatively high pressure - but nevertheless gaseous state conditions - enough of the CO₂ gas dissolves into the resin phase to have the effect of sufficiently reducing the viscosity of the resin to enable efficient mass transfer of substrate and product molecules to and from the active sites of the catalyst. Also, it is considered possible that the rate of hydrogenation is further enhanced through an increase in the solubility of H₂ gas in the hop resin acid that we suspect may occur as a consequence of the dissolution into it of CO₂.

EXAMPLE 2 (Direct Hydrogenation of Iso-α-acids under near-liquefying Pressure of CO₂ Gas)

[0025] This example illustrates that the application of a high pressure of CO₂ gas to enhance the fluidity of a hop resin acids/catalyst mixture in the working of our invention may be taken to include temperature/pressure combinations that are subcritical (i.e. < 31°C), yet close to the point at which the CO₂ will liquefy.

[0026] An autoclave was first charged with 10.34 g of Iso-α-acids resin ("Preparation 1") and 0.564 g of PdO/C catalyst ("Catalyst I "), for a catalyst to iso-α-acids loading of 3.1% on a dry weight basis). The vessel was then purged and evacuated before introducing H₂ to a pressure of 0.45 MPa (65 psig).The pressure was then raised to 5.5 MPa (88 psig) by addition of CO₂ gas. The autoclave was immersed into a water bath at 20°C and stirring commenced. After 1.7 h, the pressure had fallen to a steady value of 5.2 MPa (755 psig), at which point the stirrer was turned off, the gas pressure relieved to atmosphere and the vessel opened. A sample of the contents was taken, mixed with methanol and filtered for analysis. Results are given in Table 1. The product shows a low level of perhydrogenation (A_254nm/A_270nm = 1.20), no trace of iso-α-acids or dihydroiso-α-acids and an almost identical purity of 93.9% by HPLC.

[0027] At 20°C, carbon dioxide exists as a liquid at pressures in excess of about 5.7 MPa (828 psig) (See Handbook of Chemistry & Physics (54^th Edition) p B-94). The reaction conditions of Example 2 (wherein the partial pressure of the CO₂ gas was 5 MPa (735 psi)) were therefore approaching, though quite definitely below the point at which CO₂ liquefies. Clearly, it may be expected that the amount of CO₂ gas dissolved into the iso-α-acids resin will increase with increasing pressure. Thus, it is believed that the greater the pressure of added CO₂, the greater will be the amount of CO₂ gas dissolved, leading to a lower viscosity in the stirred resin (and a possibly faster rate of reaction) at a given temperature.

[0028] Table 1 shows that the iso-α-acids in the resin were efficiently converted to tetrahydroiso-α-acids.

Table 1: Purity of Reaction Products by HPLC and Spectrophotometry

Example No.	Purity of THIAA* (HPLC)	% DHIH**		A254nm/A270nm
1	93.6	0.0	0.90	1.19
2	93.9	0.0	0.90	1.20

DHIH = Dihydroisohumulone; THIAA = Tetrahydroiso-α-acids. * Purity of THIAA = [Area of THIAA peaks x 100]/Total peak area ** % DHIH = [Area of DHIH peak x 100]/Total peak area N/M = Not measurable due to unknown, interfering contaminants. (N.B. All HPLC peak detection was at 270nm)

Table 2: Viscosity of Iso-α-acids and Tetrahydroiso-α-acids (cP)

HOP RESIN ACID	TEMPERATURE (°C)
	40	65
Iso-α-acids	175	39
Tetrahydroiso-α-acids	410	66

[0029] In our next example, we show how our process may be applied to the hydrogenation of a β-acids resin. The purified β-acids fraction of hops is particularly viscous, being a semi-crystalline solid or extremely viscous paste at normal room temperature, and therefore cannot easily be handled unless it is substantially heated, typically to above 50°C. (See e.g. Forrest, Seaton & Moir in US Patent No. 4,717, 580).

[0030] Our next example shows how the hydrogenation of a naturally semi-solid hop resin acid may be further facilitated by the application of a high pressure of CO₂ gas at a temperature exceeding the critical temperature (31°C).

EXAMPLE 3 (Direct Hydrogenation of β-Acids Resin under High Pressure of CO₂ Gas)

[0031] A small, stainless steel autoclave (having internal volume of 885 ml) equipped with a PTFE-coated magnetic stirring bar was charged with 7.21 g of a β-acids resin containing (by HPLC) 81.0% β-acids (of which 41.9% was in the form of colupulone), 6.8% α-acids and 4.0% iso-α-acids. 0.672 g of Pd/C catalyst (Catalyst 2) was added, to give a catalyst to β-acids loading of 6.2% (dry weight basis). The autoclave was then immersed for a short time in water at 52°C to help fluidise the resin. Next, air was purged out of the vessel in the manner of Example 10 and hydrogen introduced to a pressure of 0.45 MPa (65 psig). The pressure was then increased to 5.5 MPa (800 psig) by introduction of CO₂ gas. The vessel was then partially immersed in water at 52°C for 3 min. before starting the magnetic stirrer. The temperature was maintained at 45 - 48°C for 4.0 h, at which point the stirrer was turned off and the vessel placed in a water bath at 20°C. After gently releasing the gas pressure over a period of 2 -3 min., the vessel was opened. The mixture of resinous product and catalyst was then sampled and analysed by HPLC. This analysis showed that the β-acids had almost completely disappeared (>99.5% reduction), being replaced by a complex mixture of hydrogenated products of which about 50% was in the form of hexahydro-β-acids.

[0032] Hence, it was apparent that, by virtue of the addition of a high pressure of CO₂ gas, it was possible to hydrogenate a hop resin acids preparation under relatively low temperature conditions at which the resin/catalyst mixture would otherwise have been highly viscous and difficult to stir.

[0033] In the following Examples 4A - 4F, we demonstrate that the unexpected ability of CO₂ gas to facilitate hydrogenation of hop resin acids is not merely a function of applied pressure that could be duplicated with the use of, for example, nitrogen, another, relatively inert, non-flammable gas that is commonly used to reduce or eliminate the risk of ignition of hydrogen that might leak from the reaction vessel.

EXAMPLES 4A - 4F (Direct Hydrogenation of Iso-α-acids in the Presence of Reaction Inert Gas)

[0034] In each of six different hydrogenation runs, 10.0 - 10.1 g of an iso-α-acids resin (93.8% iso-α-acids by HPLC and having peak area purity of 94.1 %, Preparation 3) was placed in the autoclave of Example 3. Next, 0.451 - 0.454 g of 10% PdO on carbon catalyst (Catalyst 1) was added and dispersed into the resin by swirling the whole vessel by hand. The vessel was then closed and purged of air by flushing with N₂ (for Examples 4E & 4F only) or CO₂, followed by application of vacuum. Hydrogen was then introduced to a final pressure of 0.4 MPa (60 psig). The temperature was maintained at 19°C by immersion in a beaker of water at that temperature. According to the nature of the particular experiment, N₂ or CO₂ gas was then admitted as (and if) required to a final pressure as indicated in Table 3. After 3 min., the magnetic stirrer was activated and the resin/catalyst mixture stirred at the same speed controller setting. After 2.0 h, the pressure was relieved to atmosphere, the vessel opened and an weighed aliquot of the contents dissolved into methanol and analysed by HPLC. Results are given in Table 3.

Table 3: Effect of Reaction Inert Gas Pressure on Hydrogenation of Iso-α-acids Resin

Example No.	Inert Gas Added	Partial Pressure of Inert Gas MPa (psi)	% Conversion of IAA to THIAA*
4A	None	0	8
4B	CO2	2.1 (300)	69
4C	CO2	3.8 (550)	100
4D	CO2	4.8 (700)	100
4E	N2	4.8 (700)	8
4F	N2	4.8 (700)	9

* % Conversion = THIAA x 100/[THIAA + IAA + DHIH]

[0035] As may be seen from Table 3, the poor hydrogenation of iso-α-acids at ambient temperature in the particular circumstances of Example No. 12A was clearly much improved when CO₂ gas was added. At a partial pressure of 2.1 MPa (300 psi) of CO₂ gas (Example 4B), the conversion of iso-α-acids to tetrahydroiso-α-acids was increased from a commercially useless value of a mere 8% to 69%, whilst at a starting value of 3.8 MPa (550 psi) (Example 4C), the reaction was actually completed within the 2 h time period, as was also the case when the partial pressure of CO₂ was set to 4.8 MPa (700psi) (Example 4D). By contrast, even the application of 4.8 MPa (700 psi) of N₂ gas (Examples 4E & 4F) failed to have any significant effect whatsoever on the reaction rate. The results of Examples 4A to 4F clearly indicate that CO₂ gas has a high solubility in the iso-α-acids resin, sufficient to be likely to have a substantial effect on the viscosity of the resin and thereby accounting, at least in part, for the considerable improvement in reaction rate that was found to occur. Furthermore, it is believed that the solubility of hydrogen into a hop resin acids/catalyst mixture may be significantly and usefully improved by the presence of a substantial quantity of dissolved CO₂ such as may occur when hydrogenation is conducted under a moderate or high (partial) pressure of this generally rather inert gas that is believed to take no direct part in the hydrogenation reaction.

[0036] The next example demonstrates that the solubility of CO₂ gas under high pressure in a hop resin acid is significant, but limited.

EXAMPLE 5 (Solubility of CO₂ Gas in Iso-α-acids Resin)

[0037] 76.8 g of an iso-α-acids resin ("Preparation 3") was placed in the autoclave of Example 3. The autoclave was partially immersed in a water bath at 19°C. Air was removed from the autoclave by purging with CO₂ and the vessel isolated. Further CO₂ gas was then slowly introduced into the autoclave, over a period of 7 minutes, until the pressure reached a value of 4.8 MPa (700 psig).The inlet valve was then closed, isolating the autoclave from the pressurised supply tank. After 3minutes, stirring of the resin was initiated using a Teflon-coated stirring bar. The temperature of the water bath was maintained at 19°C. It was observed that, after 1.3 h., the CO₂ pressure had stabilized at 4.1 MPa (601 psig), whereas in a previous, control experiment conducted in the absence of any iso-α-acids resin, the CO₂ pressure had decreased from 4.8 (700 psig) to 5 MPa (681 psig). Thus, the drop in CO₂ pressure due to dissolution of CO₂ into the iso-α-acids resin was 0.55 MPa (80 psi). Using the Ideal Gas Law equation, and making due allowance for the volume occupied by the resin, this difference in pressure drop was calculated to indicate that the amount of CO₂ dissolved into the iso-α-acids resin was 10.5% of the mass of the iso-α-acids resin.

[0038] Thus, contrary to the aforementioned teachings of Poliakoff et al and of Hill et al, we have found that hydrogenation in the presence of carbon dioxide may be conducted efficiently under circumstances wherein the CO₂ is present not as a rather poor solvent for the hop resin acids, but rather as an aid to working that enables hydrogenation of hop resin acids to take place in a highly concentrated form and at a relatively low temperature, often below the critical temperature for CO₂ or even below that at which the hop resin acid may normally be regarded as in a fluidised state, this being typically achieved by the application of a pressure of CO₂ gas that is below, or even substantially below the critical pressure for this gas.

[0039] As can be seen from the foregoing, a particular advantage of our new process permits us to conduct catalytic, heterogeneous hydrogenation of hop resin acid substrates, especially of iso-α-acids and rho-iso-α-acids, in the complete absence of conventional organic solvents, allowing the dissolution into the substrate of gaseous carbon dioxide. Surprisingly, we have found that it is possible in this way to achieve desirable and efficient conversion of substrates into hydrogenated products in high yields and at rates that are convenient for the controlled, economic operation of an industrial process.

[0040] Furthermore, as can also be seen from the foregoing, tetrahydroiso-α-acids formed from iso-α-acids by our new process when worked under appropriate conditions would be eminently suitable as a starting material for production of hexahydroiso-α-acids by means of conventional processes.

Claims

1. A process for the hydrogenation of hop resin acids in their acidic state, and in the absence of liquid, organic solvents, by contacting a reaction mixture of said hop resin acids with hydrogen gas in the presence of a noble metal catalyst under partial pressure of carbon dioxide (CO₂) the pressure of said CO₂ gas being (a) sufficient to fluidise said reaction mixture, and (b) below that at which the said CO₂ gas will liquefy at below the critical temperature for CO₂, or at pressure less than that at which the density of the CO₂ gas exceeds its value at the critical point if the temperature is at or above the said critical temperature.

2. The process of Claim 1, wherein the partial pressure of carbon dioxide gas is between 0.4 and 0.9 times the pressure at which the said CO₂ gas will liquefy at the temperature below critical temperature.

3. The process of Claim 1, wherein the temperature is between 10 and 100°C.

4. The process of Claim 1, wherein the temperature is between 15 and 50°C.

5. The process of Claim 1, wherein the hydrogen gas is supplied at or exceeding the theoretical molar ratio for complete conversion of the hop resin acid(s) to a hydrogenated product of defined molecular structure or homologous series of structures.

6. The process of Claim 5, wherein the hydrogen gas is supplied at a molar ratio of between 1.0 and 2.0 times the theoretical requirement for complete reaction.

Ansprüche

1. Verfahren zur Hydrierung von Hopfenharzsäuren in ihrer sauren Form und in Abwesenheit flüssiger organischer Lösemittel durch Zusammenbringen einer Reaktionsmischung der Hopfenharzsäuren mit Wasserstoffgas in Gegenwart eines Edelmetallkatalysators unter einem Partialdruck von Kohlendioxid (CO₂), wobei der Druck des CO₂-Gases (a) ausreicht, um das Reaktionsgemisch zu fluidisieren, und (b) sich unterhalb desjenigen Druckes befindet, bei dem sich das CO₂-Ga unterhalb der kritischen Temperatur von CO₂ verflüssigt, oder bei einem Druck, der geringer ist als derjenige, bei dem die Dichte des CO₂-Gases die Dichte am kritischen Punkt übersteigt, wenn die Temperatur bei oder oberhalb der kritischen Temperatur liegt.

2. Verfahren nach Anspruch 1, bei dem der Partialdruck des Kohlendioxidgases zwischen dem 0.4- und 0.9-fachen des Druckes liegt, bei dem sich das CO₂-Gas bei der unter der kritischen Temperatur liegenden Temperatur verflüssigt.

3. Verfahren nach Anspruch 1, bei dem die Temperatur zwischen 10 und 100° C liegt.

4. Verfahren nach Anspruch 1, bei dem die Temperatur zwischen 15 und 50° C liegt.

5. Verfahren nach Anspruch 1, bei dem das Wasserstoffgas in einer Menge zugeführt wird, die bei oder über dem theoretischen molaren Verhältnis für die vollständige Umwandlung der Hopfenharzsäure(n) in ein hydriertes Produkt definierter molekularer Struktur oder homologer Strukturreihen liegt.

6. Verfahren nach Anspruch 5, bei dem das Wasserstoffgas in einem Molverhältnis zugeführt wird, das zwischen dem 1.0- und dem 2.0-fachen der für eine vollständige Umsetzung theoretisch erforderlichen Menge liegt.

Revendications

1. Procédé d'hydrogénation d'acides résiniques de houblon dans leur état acide et en l'absence de solvants organiques liquides, par mise en contact d'un mélange réactionnel des acides résiniques de houblon avec du gaz d'hydrogène en présence d'un catalyseur de métal noble sous une pression partielle de gaz carbonique (CO₂), la pression du gaz CO₂ (a) étant suffisante pour fluidiser le mélange réactionnel et (b) se trouvant au-dessous de la pression à laquelle le gaz CO₂ se liquéfie au-dessous de la température critique de CO₂, ou à une pression qui est inférieure à celle à laquelle la densité du gaz CO₂ dépasse la densité au point critique si la température se trouve à la température critique ou au-dessus de celle-ci.

2. Procédé selon la revendication 1, dans lequel la pression partielle du gaz carbonique est entre 0,4 et 0,9 fois la pression à laquelle le gaz CO₂ se liquéfie à la température inférieure à la température critique.

3. Procédé selon la revendication 1, dans lequel la température est entre 10 et 100°C.

4. Procédé selon la revendication 1, dans lequel la température est entre 15 et 50°C.

5. Procédé selon la revendication 1, dans lequel le gaz d'hydrogène est amené en une quantité qui est à ou au-dessus d'un rapport molaire théorique pour la conversion complète de la/des acide(s) résinique(s) de houblon en un produit hydrogéné de structure moléculaire définie ou de séries de structures homologues définies.

6. Procédé selon la revendication 5, dans lequel le gaz d'hydrogène est amené dans un rapport molaire qui est entre 1,0 et 2,0 fois la quantité théorique nécessaire pour une conversion complète.